Cooling and heating

More comments on the machines we use.

Some time ago we commented on the one-sidedness of thermodynamics: hot and cold are not equal and opposite principles, as Platonic or Daoist philosophy might have it.  In fact it’s much harder to cool things down than to heat them up, which in part explains why humans mastered the use of fire long before they constructed air-conditioners.

Another effect of this one-sided relationship was brought to our attention recently.  One of our consultants had his refrigerator stop working.  It was still consuming power, the pump was still coming on, but the ice tray held only water and the milk had curdled.  Eventually the maintenance people decided it needed to be replaced, and was.

[Our writer then asked, “Why re-frigerator?  Is it assumed that everything placed therein was once cold, and had warmed?  Does no one make a frigerator?”  We try to discourage such questions, at least during working hours.]

It was only to be expected that the machine would need replacing sometime.  It produced cold by using a pump to compress the working substance, then allowing it to expand in a controlled way in another place.  Pumps have moving parts, seals fail over time, valves get worn.  But this sort of process seems so antique, so twentieth-century: producing an effect by pumping around a chemical, opening and closing valves based on a thermocouple.  It could have been designed to illustrate a thermodynamics textbook.  By contrast, the electric stove generates heat with no moving parts (well, there’s the off/on/temperature switch).  Even a gas grill is simple by comparison.  And they have correspondingly longer lives.

There are other ways to produce cold.  Our astronomer kept the detectors cold at his observatory (they need to be very cold indeed) by a constant bath of liquid nitrogen.  Here in Alexandria a large room is displayed by the historical people, in which Potomac river ice was stored during the winter to be used in the summer.  Well, this only puts the cooling back one step.  A machine is needed to make the liquid nitrogen in the first place, and winter is not a controllable quantity.

There is a solid-state device that can produce cold (or, more accurately, produce a difference in temperature).  The Peltier cooler is in fact used on some telescopes to keep the detector cool.  It doesn’t do as well as liquid nitrogen, but is much more convenient.  It is, however, far less efficient than the contraption with a pump and expansion valve, and hard to scale up.  That’s because, roughly speaking, the effects involved come from the almost-canceling of big numbers, and thus are much smaller than the numbers themselves.

Still, much of the modern world is built on small effects.  The compass, barometer and accelerometer in your smartphone all use solid-state detectors measuring tiny voltages.  It seems that someone should be able to come up with an efficient solid-state cooler.  The fact that no one has done so, when the advantages are so obvious, means it must be very difficult indeed.

Unless everyone who’s tried has realized that a refrigerator that never needed to be replaced wouldn’t bring them much money in the long run. . .